wmed2bib

#!/usr/bin/perl

use strict ;
use warnings ;

use LWP::UserAgent ;
use IPC::Open2 ;

use constant FALSE => 0 ;
use constant TRUE  => 1 ;

my $med2bib_prog = $ENV{HOME}.'/local/bin/med2bib' ;

if ( not @ARGV )
  {print STDERR "Give a list of PMID as parameter.\n" ;
   exit 1 ;
  }
elsif ( grep {not /^\d+$/} @ARGV )
  {die "PMID are numerical.\n" ;
  }

# URL documentation:
# http://www.ncbi.nlm.nih.gov/books/bv.fcgi?rid=helplinks.chapter.linkshelp#linkshelp.Retrieve_PubMed_Cita

my $pmid_list  = join ',', @ARGV ;
my $pubmed_url = join '', 'http://www.ncbi.nlm.nih.gov/pubmed/',
                          $pmid_list,
                          '?report=medline&format=text' ;

my $ua  = LWP::UserAgent->new() ;
my $req = HTTP::Request->new( GET => $pubmed_url ) ;
my $res = $ua->request( $req ) ;

# check the outcome
if (not $res->is_success() )
  {die "Error: " . $res->status_line() . "\n" ;
  }


my $pubmed_records = $res->decoded_content() ;
$pubmed_records =~s/\A.*<pre>\n//s ;
$pubmed_records =~s/<\/pre>.*\z/\n\n\n/s ;

#print $pubmed_records ;


foreach my $record ( split /\n\n+/, $pubmed_records )
  {my ( $prg_out, $prg_in );
   my $pid = open2($prg_out, $prg_in,
                   $med2bib_prog ) or die "Cannot open $med2bib_prog" ;

   print $prg_in $record ;
   close $prg_in ;

   my $output = '' ;
   my $is_inside = FALSE ;
   my $in_title = FALSE ;
   my $pmid ;
   while ( my $line = <$prg_out> )
     {# On reformate la ligne du PMID pour supprimer les espaces en trop
      $line =~s/^line\s*=\s*PMID\s*-\s*(\d+)/line=PMID-$1/m ;

      if ( defined $1 )
	{$pmid = $1 ;
	}

      # On copie le PMID dans le corps de la fiche:
      if ( $is_inside and defined $pmid )
	{$output .= qq#note    = \{PMID $pmid\},\n# ;
         undef $pmid ;
	}

      # On transforme les traits d'unions en tirets (pour l'anglais)
      #$line =~s/(pages\s*=\{[0-9A-Za-z]+)\-([0-9A-Za-z]+\},)/$1--$2/gm ;

#      # Dans le titre, on met les mots contenant des majuscules entre accolades
#      if ( $line =~/^title\s*=/ )
#	{$in_title = TRUE ;
#        }
#      if ($in_title)
#	{while( $line =~/(\S+)/g )
#	   {my $word = 
#	   }
#        }
#      if ( $in_title and $line =~/\},\s*$/ )
#	{$in_title = FALSE ;
#	}


      $output .= $line ;

      if ( $line =~m/^\@/ )
	{$is_inside = TRUE ;
	}
    }


   print STDOUT $output ;
  }


exit 0 ;